How it works

What a PlugVet grade means, in five steps. The rules themselves live on How we grade.

1. One public commit

PlugVet reads the public source of a skill, plugin or MCP server at one commit, and records the date. The grade describes that commit on that date.

2. Eight automated checks

Permissions, Install and update, Network, Secrets, Agent instructions, Obfuscation, Licence and Maintainer activity. What each check looks for, and what it does not.

3. Two scales

Each check gets a check outcome, such as Pass or Note. Each finding gets a finding severity: Note, Concern or Serious. The two scales.

4. The findings decide the letter

  • A, No findings from our checks
  • B, Minor notes only
  • C, One Concern
  • D, Two or more Concerns, or a Serious

How the letter is chosen.

5. What you can do next

  • Read the evidence. Every finding on an entry page names a file and a line.
  • Watch a tool. Alert emails aren't switched on yet. Watch alerts
  • Challenge a grade. Nicholas Wells checks and approves each challenge and correction.

A worked example

Rails Consultant (thoughtbot/rails-consultant) is graded A, No findings from our checks. All eight checks passed at commit 557f1b07f4c2. Checked with PlugVet grader 2h2 on 14/11/2026.

A hygiene check of the source at the commit shown, not a security audit or a guarantee.

Limits

Grades are a hygiene check of the source at the commit shown, on the date checked, not a security audit or a guarantee. What a grade is not.

AI built the scanner and checks. Nicholas Wells checks and approves each paid report and grade challenge. Free directory grades, A to D, are automated and not checked one by one.